Integration of Virtual Reality and Animation Design Character Action Interaction Technology Based on Computer Intelligence

This paper studies the integration of virtual reality (VR) based on computer intelligence and animation design character action interaction technology, aiming to solve the current problems of unnatural virtual character action generation, high interactive response delay and insufficient user immersion. Existing technologies have limitations in processing complex actions and real-time interactions, resulting in poor user experience. To achieve the goal, this paper first uses the deep learning model LSTM to train large-scale motion capture data to build a high-precision motion library; then uses the reinforcement learning algorithm PPO to optimize the character's real-time decision-making ability in the virtual environment; finally, combines the multimodal sensor inertial measurement unit and optical tracking system to achieve high-precision capture and mapping of user movements, and reduces motion delay through an adaptive algorithm. Experiments show that this method can achieve an action generation speed of 108.8 FPS and reduce the action delay to less than 27 ms, which is significantly better than traditional methods. This technology effectively improves the naturalness of virtual character actions and the efficiency of interactive response, and provides an innovative solution for the fields of VR and animation design.